CA2956107A1 - Improved host cell for producing proteins - Google Patents

Abstract

The present invention relates to a host cell for producing a recombinant peptide, polypeptide or protein of interest, wherein said host cell comprises at least 2 copies of a nucleic acid sequence encoding a poison protein; and to the use thereof for producing peptides, polypeptides or proteins of interest.

Description

IMPROVED HOST CELL FOR PRODUCING PROTEINS
FIELD OF INVENTION
The present invention relates to heterologous systems for recombinant protein production.
In particular, the present invention relates to host cells for use in methods for producing proteins based on an improved poison-antidote system.
BACKGROUND OF INVENTION
Microorganisms are today widely used for protein or DNA production. Usually, these processes require the use of bacterial plasmids as vectors carrying the gene to be expressed. However, plasmid maintenance within microorganisms is limited, and plasmid instability thus represents a significant concern in recombinant protein production or in DNA production.
It has been demonstrated that the growth rate of plasmid-bearing cells is significantly reduced relative to that of a plasmid-free cell. One theory is that plasmid replication and transcription, as well as protein production, represent a significant burden on cellular metabolism. Hence, in a fermentation process, cells losing the plasmid exhibit a higher fitness than cells still bearing the plasmid, and the former rapidly overcome the latter in the bacterial population.
In order to limit plasmid loss in a cell population, and to avoid plasmid-free cells to survive and dominate the culture, selectable markers were inserted in plasmids.
The most common selectable markers used in fermentation procedures are antibiotic-resistance genes. However, contamination of the product or biomass by antibiotics (or genes encoding an antibiotic resistance) is unacceptable from a medical or regulatory perspective. Moreover, antibiotic-resistance genes may propagate in the environment, or be transferred to pathogenic strains. Moreover, recent studies demonstrated that using

2 antibiotic-resistance genes for producing a recombinant protein strongly reduces the yield of protein production (Peubez et al., Microbial Cell factories, 2010, 9:65).
There is thus a need for other systems allowing plasmid maintenance, and free from drawbacks of antibiotic-resistance genes.
An alternative to the use of antibiotic-resistance genes is the complementation of an essential mutated chromosomal gene by a wild-type allele inserted into the plasmid. For example, systems were developed in which the mutant host is unable to synthesize an essential amino acid without a plasmid carrying out the gene that provides this function.
However, this approach seriously restricts the possible choices in growth medium.
Another strategy developed is a system in which a plasmid-mediated repressor titration overcomes the repression of an essential chromosomal gene placed under the control of the lac operator. However, this procedure has the following limitations: (i) it makes the Lac promoter unavailable for other purposes such as protein expression, (ii) the system is limited to E. coli or other bacteria wherein the Lac promoter is functional, and (iii) medium containing lactose has to be avoided.
Another alternative system to the use of antibiotic-resistance genes is based on couples of poison proteins (i.e. molecules which are toxic for the host cell) combined to their antidotes. For example, the poison gene may be expressed by the host cell from a chromosomal copy, while the antidote is carried by the plasmid. Therefore, presence of the plasmid is required for the host cell survival. An example of couple poison/antidote is the ccdA (antidote) / ccdB (poison) system. CcdA and CcdB are the antidote and toxin proteins encoded by the E. coli F plasmid. Together, they ensure the death of daughter cells that do not receive a copy of F. Expression of the ccdB protein interferes with the rejoining step of DNA gyrase, causing the host cell chromosome to be cut to pieces. Other examples of antidote / poison couples include, but are not limited to, Kis/Kid proteins, Phd/Doc proteins, RelB/relE proteins, PasB (or PasC)/PasA proteins, mazF/maze proteins.
This antidote/poison system has been extensively used in cloning methods, as described for example in the US patent US8,470,580: a host cell comprising the gene encoding a

3 poison is used in combination with a plasmid comprising the gene encoding the corresponding antidote. This system thus allows the direct selection of cells having integrated a gene of interest, as only cells expressing the antidote gene survive.
This system is also used for protein production, as previously described by the inventors (Szpirer and Milinkovitch, BioTechniques, 2005, 38(5):775-781). Unexpectedly, an increase of three to five-fold of the recombinant protein production level was observed, demonstrating the great potential of this system for producing recombinant proteins.
However, the Inventors observed that, when this system is used for producing toxic proteins, some drawbacks may appear, such as, for example, mutations within the gene of interest, mutations in the antidote/poison system and the like. Willing to develop an improved plasmid stabilization system, they modified the classic system comprising one copy of the antidote gene combined with one copy of the poison gene.
Surprisingly and unexpectedly, they showed that the insertion of an additional copy of the poison gene in the genome of the host cell increases the stability of the plasmid, but also increases the yield of protein production (in a non-correlate manner).
SUMMARY
The present invention thus relates to a host cell for producing a recombinant peptide, polypeptide or protein of interest, wherein said host cell comprises at least 2 copies of a nucleic acid sequence encoding a poison protein. In one embodiment, said at least 2 copies are in a different replicon than a nucleic acid sequence encoding the antidote protein to the poison protein In one embodiment, said poison protein is CcdB, encoded by SEQ ID NO: 1 or any nucleic acid sequence having at least 75% identity with SEQ ID
NO: 1.
In one embodiment, the host cell further comprises at least one copy of a nucleic acid sequence encoding the antidote protein to the poison protein. In one embodiment, said antidote protein is CcdA, encoded by SEQ ID NO: 13 or any nucleic acid sequence having at least 75% identity with SEQ ID NO: 13.

4 In one embodiment, the nucleic acid sequence encoding the antidote protein is carried by a plasmid further comprising an expression system wherein the nucleic acid sequence encoding a recombinant peptide, polypeptide or protein of interest is or may be inserted.
In one embodiment, said expression system comprises a promoter selected from the group comprising a T7 promoter, Ptrc, Para and Plac. In another embodiment, said expression system comprises a T7 promoter and the host cell further comprises a T7 RNA
polymerase gene, preferably inserted in its genome, more preferably in a T7 expression system. In another embodiment, the expression system comprises a T7 promoter and the host cell further comprises a genetically modified phage inserted within its genome, preferably said phage is defined as a phage wherein:
- a T7 expression system is inserted, - the S, R, and/or the Q genes are inactivated, and - the Int and/or Xis gene are inactivated.
In one embodiment, the host cell is a bacterium, preferably a gram negative bacterium, more preferably an Enterobacteriacea, and even more preferably E. coli. In one embodiment, the host cell of the invention further comprises inactivation of at least one of the genes tonA, galK, araB, araA, Ion, ompT, rcsA, hsdR, mrr, endA and recA. In one embodiment, the host cell of the invention further comprises at least one additional copy of the gyrA gene.
The present invention also relates to a kit comprising a host cell as described hereinabove and a vector comprising at least one copy of the nucleic acid sequence encoding the antidote protein and at least one copy of an expression system wherein the nucleic acid sequence encoding a recombinant peptide, polypeptide or protein of interest is or may be inserted. In one embodiment, the nucleic acid sequence encoding a peptide, polypeptide or protein of interest is under the control of a promoter selected from the group comprising a T7 promoter, Ptrc, Para and Plac.
Another object of the invention is a method for producing a recombinant peptide, polypeptide or protein of interest, wherein said method comprises cultivating the host cell as described hereinabove and recovering the peptide, polypeptide or protein of interest.

In one embodiment, the recombinant protein is a secreted protein, a transmembrane protein or a protein which is toxic for the bacterial strain.
DEFINITIONS

5 In the present invention, the following terms have the following meanings:
- As used herein, a "peptide" refers to a linear polymer of amino acids of less than 50 amino acids linked together by peptide bonds; a "polypeptide" refers to a linear polymer of at least 50 amino acids linked together by peptide bonds; and a protein specifically refers to a functional entity formed of one or more peptides or polypeptides, and optionally of non-polypeptides cofactors.
- "Recombinant peptide, polypeptide or protein" refers to a peptide, polypeptide or protein generated from recombinant DNA, i.e. from DNA artificially inserted in a producing host cell.
- "Poison protein" refers to a protein which is toxic for the host cell producing it. As used herein, a poison protein is thus toxic for the host cell of the invention.
DETAILED DESCRIPTION
The present invention relates to a host cell for producing recombinant peptides, polypeptides or proteins, wherein said host cell comprises at least two copies of a nucleic acid sequence encoding a poison protein. In one embodiment, the host cell comprises 2 copies of the nucleic acid sequence encoding the poison protein, or 3, 4, 5 or

6 copies (or more) of this nucleic acid sequence.
Examples of poison proteins include, but are not limited to, CcdB, Kid, Doc, RelE, PasA, MazE or any other poison molecule such as for example bacteriocins which is or is not of plasmid origin. The poison protein can also be a toxin protein being naturally or artificially toxic and affecting one or more vital functions of a (prokaryote) cell. The protein encoded by the gene sacB (from Bacillus amylolique-faciens), the protein GpE, the protein GATA-1 and the protein Crp are other examples of such toxic molecules. The

7 gene sacB encodes a levan sucrase which catalyses the hydrolysis of sucrose into products which are toxic for E. Coll (Pierce et al. Proc. Natl. Acad. Sci., Vol. 89, N[deg16 (1992) p. 2056-2060). The protein GpE is encoded by the E genes from the bacteriophage [phi]X174 which includes six unique restriction sites. GpE causes lysis of E.
Coll cell (Heinrich et al., Gene, Vol. 42(3) (1986) p. 345-349). The protein GATA-1 has been described by Trudel et al. (Biotechniques 1996, Vol. 20(4), p. 684-693). The protein Crp has been described by Schlieper et al. (Anal. Biochem. 1998, Vol. 257(2), p.
203-209).
Preferably, the host cell of the invention comprises at least two, preferably 2, copies of a nucleic acid sequence encoding the protein CcdB. In one embodiment, the nucleic acid sequence encoding the protein CcdB is the ccdB gene (sequence SEQ ID NO: 1) or any sequence encoding a functional CcdB protein having at least 75%, preferably at least 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% or more identity with SEQ ID NO:
1.
In one embodiment, the at least 2 copies of a nucleic acid sequence encoding a poison protein (preferably CcdB) have not 100% identity (for example have about 30%, preferably about 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% or more identity), in order to avoid recombination events between 2 copies.
In one embodiment, the at least 2 copies of a nucleic acid sequence encoding a poison protein (preferably CcdB) are from different origins (for example are from at least 2 bacterial species) or the first one is of plasmid origin and the second one is of chromosomal origin. For example, the first copy of CcdB may be from E. coli 0157:H7 and the second copy of CcdB may be from the F plasmid.
In one embodiment, the at least two poison proteins are 100% identical, but are encoded by nucleic acid sequences that do not have 100% identity (i.e. the nucleic acid sequences differ by non-coding differences, such as, for example, a codon encoding a given amino acid is replaced by another codon encoding the same amino acid).
In one embodiment, the at least two poison proteins are not 100% identical (for example have about 30%, preferably about 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94,95, 96, 97, 98, 99% or more identity), but their effect may be counteracted by the same antidote.

In one embodiment, the at least 2 copies of a nucleic acid sequence encoding a poison protein (preferably CcdB) have not 100% identity, but both contain conserved putative domains according to in silico analysis, using for example, the program TAQ
V1.0 described in Guglielmini et al. (BMC Microbiology 2008, 8:104). In one embodiment, the second copy of a nucleic acid sequence encoding a poison protein is a putative poison protein related to the first one and identified using the program TAQ V1Ø In one embodiment, the first copy of a nucleic acid sequence encoding a poison protein is CcdB, and the second copy is selected among the 22 CcdB in silico inferred toxins identified using the program TAQ V1.0 by Guglielmini etal., 2008.
In one embodiment of the invention, the at least two copies of a nucleic acid sequence encoding a poison protein (preferably CcdB) are on a different replicon than the nucleic acid sequence encoding the corresponding antidote protein, i.e. on a replicon that does not comprise a nucleic acid sequence encoding the corresponding antidote protein. In particular, according to this embodiment, host cells comprising at least two copies of the plasmid F (plasmid F comprises a replicon comprising both a nucleic acid sequence encoding the poison protein and a nucleic acid sequence encoding the antidote protein) are not part of the present invention.
In one embodiment, the at least two copies of a nucleic acid sequence encoding a poison protein (preferably CcdB) are on the same replicon. In another embodiment, the at least two copies of a nucleic acid sequence encoding a poison protein (preferably CcdB) are on at least two distinct replicons.
In one embodiment, at least one of the at least two copies of a nucleic acid sequence encoding a poison protein (preferably CcdB) is on a plasmid. In another embodiment, the at least two copies of a nucleic acid sequence encoding a poison protein (preferably CcdB) are on a plasmid.
In one embodiment of the invention, the at least two copies of a nucleic acid sequence encoding a poison protein (preferably CcdB) are chromosomal, i.e. are inserted within the genome of the host cell.

8 In one embodiment of the invention, when two copies of the nucleic acid sequence encoding a poison protein are inserted within the genome of the host cell, each stands at an opposite side on the chromosome of the host cell. Preferably, each copy is inserted in a non-coding region of the chromosome of the host cell or in a gene to be inactivated.
For example, a first copy of a nucleic acid sequence encoding a poison protein is inserted in the dcm gene, and a second copy of a nucleic acid sequence encoding a poison protein is inserted between the yjjK and sit genes, i.e. at the opposite side of the E. coli chromosome, in an intergenic non-coding space.
The term "identity" or "identical", when used in a relationship between the sequences of two or more nucleic acid sequences or polypeptides, refers to the degree of sequence relatedness between nucleic acid sequences or polypeptides (respectively), as determined by the number of matches between strings of two or more nucleic acid residues or amino acid residues (respectively). "Identity" measures the percent of identical matches between the smaller of two or more sequences with gap alignments (if any) addressed by a particular mathematical model or computer program (i.e., "algorithms").
Identity of related polypeptides can be readily calculated by known methods. Such methods include, but are not limited to, those described in Computational Molecular Biology, Lesk, A. M., ed., Oxford University Press, New York, 1988; Biocomputing: Informatics 5 and Genome Projects, Smith, D. W., ed., Academic Press, New York, 1993; Computer Analysis of Sequence Data, Part 1, Griffin, A. M., and Griffin, H. G., eds., Humana Press, New Jersey, 1994; Sequence Analysis in Molecular Biology, von Heinje, G., Academic Press, 1987;
Sequence Analysis Primer, Gribskov, M. and Devereux, J., eds., M. Stockton Press, New York, 1991; and Carillo et al., SIAM J. Applied Math. 48, 1073 (1988).
Preferred methods for determining identity are designed to give the largest match between the sequences tested. Methods of determining identity are described in publicly available computer programs. Preferred computer program methods for determining identity between two sequences include the GCG program package, including GAP (Devereux et al., Nucl.
Acid. Res. \2, 387 (1984); Genetics Computer Group, University of Wisconsin, Madison, Wis.), BLASTP, BLASTN, and FASTA (Altschul et al., J. MoI. Biol. 215, 403-410 (1990)). The BLASTX program is publicly available from the National Center for

9 Biotechnology Information (NCBI) and other sources (BLAST Manual, Altschul et al.
NCB/NLM/NIH Bethesda, Md. 20894; Altschul et al., supra). The well-known Smith Waterman algorithm may also be used to determine identity.
Preferably, the nucleic acid sequence encoding the poison protein is under the control of the native promoter. For example, the nucleic acid sequence encoding the protein CcdB, may be fused downstream to the ccd promoter. Therefore, in the presence of the antidote, the ccd promoter is strongly repressed.
In an embodiment, the host cell is a microorganism, preferably a prokaryote, more preferably a bacterium, and even more preferably a gram negative bacterium.
Advantageously, the host cell is a bacterium from the Enterobacteriacea family according to the current applicable taxonomy. Should the taxonomy change, the skilled artisan knows how to adapt the changes in the taxonomy to deduce the strains that could be used in the present invention. Examples of bacteria from the Enterobacteriacea family include, but are not limited to, bacteria belonging to the genera Escherichia, Enterobacter, Erwinia, Klebsiella, Pantoea, Photorhabdus, Providencia, Salmonella, Serratia, Shigella, Morganella and Yersinia. According to a preferred embodiment, the host cell belongs to the Escherichia genus, and more preferably the host cell is Escherichia coli (E. coli). Examples of strains of E. coli which could be used in the present invention include, but are not limited to, strains derived from E. coli K-12, E. coli B
or E. coli W, such as, for example, MG1655, W3110, DG1, DG2, Top10, DH10B, DH5alpha, HM5174, BL21, BL21(DE3), HM5174(DE3), BL21(DE3) pLysS and BL21(DE3) pLysE.
In one embodiment, genes of the host cells may be inactivated.
As used herein, the term "inactivated" refers to the interruption or to the suppression of the expression of a gene at transcriptional or translational levels.
Preferably, the term "inactivated" refers to a gene whose transcription is suppressed. According to the invention, the inactivation of a gene may be due to the mutation of the gene or to the insertion of an expression system within the coding sequence of the gene. In the meaning of the present invention, the term "mutation" refers to a stable change in the genetic sequence. Examples of mutation which could lead to the inactivation of a gene in the present invention include, but are not limited to, point mutations, insertions, deletions and amplification or gene duplication. Preferably, the mutation is a deletion. The term "deletion" as used herein means the loss or absence of a gene, preferably the total loss or 5 absence of a gene. More preferably, the deletion starts at or before the start codon of the deleted gene, and ends at or after the stop codon of the deleted gene.
In one embodiment, the gene tonA (also known as fhuA, SEQ ID NO: 2) is inactivated.
The TonA/FhuA protein is a receptor for the phages Ti, T5 and Phi80.
In one embodiment, the gene galK (SEQ ID NO: 3) is inactivated. The deletion of this

10 gene allows the use of the galK positive/negative selection for deletion of genes by a method based on homologous recombination.
In one embodiment, the gene araB (SEQ ID NO: 4) is inactivated. In another embodiment, the gene araA (SEQ ID NO: 5) is inactivated. The inactivation of araB
and/or araA is recommended for the use of the Para promoter (inducible by arabinose) within the host cell.
In one embodiment, the gene Ion (SEQ ID NO: 6) and/or the gene ompT (SEQ ID
NO: 7) are inactivated. The Lon protein is an ATP dependent protease. The OmpT
protein is an outer membrane protease. Preferably, the genes Ion and ompT are inactivated.
In one embodiment, the gene rcsA (SEQ ID NO: 8) is inactivated. The protein RcsA is a positive regulator of the synthesis of the capsule, which is degraded by the Lon protease.
In one embodiment, the gene hsdR (SEQ ID NO: 9) and/or the gene mrr (SEQ ID
NO: 10) are inactivated. The HsdR and Mrr proteins are restriction enzymes with different specificity. Preferably, the genes hsdR and mrr are both inactivated.
In one embodiment, the gene endA (SEQ lD NO: 11) and/or the gene recA (SEQ lD
NO: 12) are inactivated. EndA is a DNA specific endonuclease. RecA is a recombination protein with protease and nuclease activity. Preferably, the genes endA and recA are both inactivated.

11 In one embodiment of the invention, at least one of the genes tonA, galK, araB, araA, Ion, ompT, rcsA, hsdR, mrr, endA and recA are inactivated. Preferably, the inactivated genes are deleted.
In a preferred embodiment, the genes tonA, galK, araB, Ion, ompT, rcsA, hsdR, mrr, endA
and recA are inactivated. Preferably, the genes tonA, galK, araB, Ion, ompT, rcsA, hsdR, mrr, endA and recA are deleted.
In one embodiment, the host cell of the invention comprises at least one additional copy of the nucleic acid sequence encoding the target of the poison protein. In one embodiment, the host cell comprises 2 copies of the nucleic acid sequence encoding the target of the poison protein, or 3, 4, 5 or 6 copies (or more) of this nucleic acid sequence. Preferably, said at least one additional copy is chromosomal, i.e. is inserted within the genome of the host cell.
In one embodiment where the poison protein is CcdB, the host cell of the invention may comprise at least one additional copy of the nucleic acid sequence encoding the protein GyrA.
In one embodiment, the nucleic acid sequence encoding the protein GyrA is the gyrA
gene (sequence SEQ ID NO: 15) or any sequence encoding a functional GyrA
protein having at least 75%, preferably at least 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% or more identity with SEQ ID NO: 15.
In one embodiment, the second copy of the nucleic acid sequence encoding the target protein of the poison protein is inserted in the chromosome of the host cell at the same location that a copy of the nucleic acid sequence encoding the poison protein.
In one embodiment, the second copy of the nucleic acid sequence encoding the target protein of the poison protein is inserted concomitantly to a copy of the nucleic acid sequence encoding the poison protein.
In one embodiment, the host cell of the invention comprises at least one copy of a nucleic acid sequence encoding the antidote protein counteracting the effects of the poison protein expressed by the host cell of the invention.

12 Examples of antidote proteins include, but are not limited to, CcdA (antidote of CcdB), Kis (antidote of Kid), Phd (antidote of Doc), RelB (antidote of RelE), PasB
(or PasC) (antidote of PasA), MazF (antidote of MazE) or immunity molecules (antidotes of bacteriocins). The antidote protein to a toxic molecule is any protein able to reduce or suppress the effect of the corresponding toxic molecule on a cell (preferably a prokaryotic cell), when said toxic molecule is produced by said cell.
According to a preferred embodiment, the poison protein is CcdB and the host cell comprises at least one copy of a nucleic acid sequence encoding CcdA. In one embodiment, the nucleic acid sequence encoding the protein CcdA is the ccdA
gene (sequence SEQ ID NO: 13) or any sequence encoding a functional CcdA protein having at least 75%, preferably at least 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% or more identity with SEQ ID NO: 13. As used herein, a "functional" CcdA protein relates to a protein having at least 75% identity with SEQ ID NO: 13 and retaining the ability of CcdA to reduce or suppress the effect of CcdB on a cell.
In one embodiment, the nucleotide sequence encoding the antidote protein is carried by a vector, such as, for example, a plasmid, a bacteriophage, a virus, a cationic vesicle or any other type of vector, preferably by a plasmid. Therefore, the skilled artisan may easily deduce that the number of copies of the nucleic acid sequence encoding the antidote protein depend on the type of plasmid used. For example, a plasmid of the pBR322 family is usually present in a cell in a number of copies ranging from about 1 to about 100, preferably from about 20 to about 50; while a plasmid of the pUC family is usually present in a cell in a number of copies ranging from about 100 to about 500, preferably from about 150 to about 200. However, the skilled artisan may also easily deduce that the number of copies of a plasmid within a host cell may also depend on the replication rate and/or on the growth of the host cell.
In a preferred embodiment, the host cell of the invention thus further comprises at least one copy of a vector, preferably a plasmid, carrying (1) a nucleic acid sequence encoding the antidote to the poison protein, and (2) a nucleic acid sequence encoding the recombinant peptide, polypeptide or protein of interest to be produced or an insertion site

13 for inserting the nucleic acid sequence encoding the recombinant peptide, polypeptide or protein of interest to be produced.
In one embodiment, the nucleic acid sequence encoding the recombinant peptide, polypeptide or protein of interest is comprised within an expression system;
or the insertion site is such that, when inserted, the nucleic acid sequence encoding the recombinant peptide, polypeptide or protein of interest is comprised within an expression system. In particular, the nucleic acid sequence encoding the recombinant peptide, polypeptide or protein of interest is placed, or will be placed after insertion, under the control of a promoter.
As used herein, an "expression system" refers to a linear or a circular DNA
molecule composed of a fragment encoding the nucleic acid sequence encoding the recombinant peptide, polypeptide or protein of interest operably linked to an additional fragment for the transcription of the system.
The additional fragment includes a promoter and a stop codon sequence. The expression system may further contain one or more origins of replication, one or more selection markers and a sequence encoding a ribosome binding site.
"Operably linked" means that fragments are arranged to be functioning as they are supposed to be, for example once transcription starts at the promoter, it goes through coded fragment to stop codon.
"Promoter" in the meaning of the present invention is an expression control element that permits binding of RNA polymerase and the initiation of transcription.
In one embodiment of the invention, the nucleic acid sequence is under the control of a "strong" promoter. A strong promoter is characterized by a high binding affinity of the promoter sequence to an RNA polymerase, usually the naturally occurring corresponding RNA polymerase, on the one hand and by a high rate of formation of mRNA by that RNA
polymerase on the other hand.
In a preferred embodiment, the nucleic acid sequence is under the control of an "inducible promoter". An "inducible promote?' is a promoter that may be regulated by external

14 factors, e.g. the presence of an inductor (also termed "inducer") molecule or the absence of a repressor molecule, or physical factors like increased or decreased temperature, osmolarity, or pH value. Different promoters and the respective induction principles were reviewed by Makrides et al. (Microbiological Reviews, 1996, (60)3: 512-538).
Examples of inducible promoters which may be used in the present invention include, but are not limited to, the tac or the trc promoter, the lac or the lacUV5 promoter (all inducible by lactose or its analog lPTG (isopropylthiol-B-D-galactoside)), the tightly regulatable araBAD promoter (Para; Guzman et al., 1995, inducible by arabinose), the trp promoter (inducible by B-indole acrylic acid addition or tryptophan starvation, repressible by tryptophan addition), the lambda promoter pL (X) (induction by an increase of temperature), the phoA promoter (inducible by phosphate starvation), the PprpB

(induction with propionate) or other promoters suitable for recombinant peptide, polypeptide or protein expression, which all use E. coli RNA polymerase.
Among inducible promoters are those that show a "leaky" expression behavior.
Such promoters (so-called "leaky promoters") are, in principle, inducible, but show nevertheless also basal expression without being externally induced. Inducible promoters that show leaky expression under non-induced conditions may behave similarly to constitutive promoters (i.e. they are steadily and continuously active or they may be activated or enhanced as a result of certain cultivation conditions). Leaky promoters may be particularly useful for continuously operated cultivation processes.
Examples of leaky promoters are the T7 promoter and the trp promoter. In the meaning of the present invention, the term "T7 promoter" includes promoters that are present in the genome of bacteriophage T7, as well as consensus sequences and variants of such promoters with the ability to mediate transcription by the T7 RNA polymerase. The bacteriophage T7 contains seventeen different promoter sequences, all of which comprise a highly conserved nucleotide sequence.
In one embodiment of the invention, the promoter may also be constitutive, i.e. a promoter which controls expression without the need for induction on the one hand, or the possibility of repression on the other hand. Hence, there is continuous and steady expression at a certain level. As an example, the strong constitutive HCD
promoter (Poo et al., Biotechnology Letters, 2002, 24:1185-1189; Jeong et al., Protein expression and purification, 2004, 36:150-156) may be applied for constitutive expression.
Advantageously, the expression of the peptide, polypeptide or protein of interest is induced in particular conditions, such as, for example, under selection.
5 According to a preferred embodiment, the nucleic acid sequence encoding the peptide, polypeptide or protein of interest is placed under the control of a T7 promoter, of a promoter inducible by IPTG, such as, for example, a Ptrc promoter or a Ptac promoter, or of a promoter inducible by Arabinose, such as, for example, Para.
In a preferred embodiment, the nucleic acid sequence encoding the antidote protein is 10 directly functional, i.e. it directly encodes a functional antidote protein. Moreover, in one embodiment, the nucleic acid sequence encoding the peptide, polypeptide or protein of interest or the insertion site is not placed within the nucleic acid sequence encoding the antidote protein.
In one embodiment, the vector of the invention comprises a cer locus. The cer locus

15 allows stable inheritence of ColE1 and related plasmids by preventing the runaway accumulation of multimers known as "dimer catastrophe". Multimer resolution is achieved through action of the XerCd site-specific recombinase at the cer site. Preferably, when a cer locus is inserted within the vector of the invention, the host cell is Escherichia coli.
In one embodiment, the vector of the invention does not comprise any antibiotic resistance gene. In another embodiment, the host cell of the invention does not comprise any antibiotic resistance genes. In one embodiment, neither the vector nor the host cell of the invention comprise any antibiotic resistance gene. Therefore, according to this embodiment, no antibiotic is required when producing peptide, polypeptide or protein of interest using the host cell and the vector of the invention.
The present invention also relates to a kit comprising a host cell of the invention, comprising at least two copies of a nucleic acid encoding a poison protein, and at least

16 one vector (preferably a plasmid) as described hereinabove. Preferably, said vector carries at least one nucleic acid sequence encoding the antidote protein to the poison protein and a nucleic acid sequence encoding the recombinant peptide, polypeptide or protein of interest in an expression system, or an insertion site for inserting the nucleic acid sequence encoding the recombinant peptide, polypeptide or protein of interest in an expression system.
According to a preferred embodiment, the host cell of the invention comprises at least two nucleic acid sequences encoding the protein CcdB inserted in its genome, and the vector of the invention carries at least one nucleic acid sequence encoding CcdA.
In one embodiment, the vector comprises an expression system and an insertion site for inserting the nucleic acid sequence encoding the peptide, polypeptide or protein of interest within said expression system, thereby obtaining a functional vector for producing the peptide, polypeptide or protein of interest. Therefore, prior to transforming host cells with the vector of the invention for producing the peptide, polypeptide or protein of interest, the nucleic acid sequence encoding the peptide, polypeptide or protein of interest has to be inserted within the vector at the insertion site.
Methods for inserting a nucleic acid sequence within an insertion site are well-known to the skilled artisan, and are usually based on the use of restriction enzymes.
With regard to the peptide, polypeptide or protein of interest, it may refer to any peptide, polypeptide or protein that is to be produced on a manufacturing scale, e.g.
an industrial peptide, polypeptide or protein or a therapeutic peptide, polypeptide or protein.
Examples for peptide, polypeptide or protein that can be produced by the method of the invention are, without limitation, enzymes, regulatory proteins, receptors, peptides (e.g.
peptide hormones), cytoldnes, antibodies, nanobodies, membrane or transport proteins.
The peptide, polypeptide or protein of interest may also be antigens as used for vaccination, vaccines, antigen-binding proteins, immune stimulatory proteins, allergens, full-length antibodies or antibody fragments or derivatives. Antibody derivatives may be selected from the group of single chain antibodies, (scFv), Fab fragments, F(ab')2

17 fragments, Fv fragments, single domain antibodies (VH or VL fragment), camelid single variable domains (VHH) or other antibody formats as described for instance in Andersen and Reilly (Current Opinion in Biotechnology, 2004, 15:456-462) or Holliger and Hudson (Nature Biotechnology, 2005 (23)9: 1126-1136).
The peptide, polypeptide or protein of interest in the present invention can also be exemplified by protein (viral antigen), e.g., coat protein, core protein, protease, reverse transcriptase, integrase, and so forth, encoded in the genome of a pathogenic virus, e.g., hepatitis B virus, hepatitis C virus, I-HV, influenza, and so forth; growth factors such as platelet-derived growth factor (PDGF), stem cell growth factor (SCF), hepatocyte growth factor (HGF), transforming growth factor (TGF), nerve growth factor (NGF), epidermal growth factor (EGF), fibroblast growth factor (FGF), insulin-like growth factor (IGF), and so forth; cytokines such as tumor necrosis factor, interferon, interleukin, and so forth;
hematopoietic factors such as erythropoietin, granulocyte colony-stimulating factor, granulocyte-macrophage colony-stimulating factor, macrophage colony-stimulating factor, thrombopoietin, and so forth; peptide hormones such as luteinizing hormone-releasing hormone (LB-RH), thyrotropin-releasing hormone (TRH), insulin, somatostatin, growth hormone, prolactin, adrenocorticotropic hormone (ACTH), melanocyte-stimulating hormone (MSH), thyroid-stimulating hormone (TSH), luteinizing hormone (LU), follicle-stimulating hormone (FSH), vasopressin, oxytoxin, calcitonin, parathyroid hormone (PTH), glucagon, gastrin, secretin, pancreozymin, cholecystokinin, angiotensin, human placenta lactogen, human chorionic gonadotropin (HCG), cerulein, motilin, and so forth; analgesic peptides such as enkephalin, endorphin, dynorphin, kyotorphin, and so forth; enzymes such as superoxide dismutase (SOD), uroldnase, tissue plasminogen activator (TPA), asparaginase, kallikrein, and so forth;
peptide neurotransmitters such as bombesin, neutrotensin, bradykinin, substance P.
Alzheimer's amyloid peptide (AD), SOD1, presenillin 1 and 2, renin, Dsynuclein, amyloid A, amyloid P. activin, anti-HER-2, bombesin, enkephalinase, protease inhibitors, therapeutic enzymes, D 1- antitrypsin, mammalian trypsin inhibitor, mammalian pancreatic trypsin inhibitor, calcitonin, cardiac hypertrophy factor, cardiotrophins (such as cardiotrophin-1), CD proteins (such as CD-3, CD-4, CD-8 and CD-19), CFTR, CTNF, DNase, human chorionic gonadotropin, mouse gonadotropin-associated peptide,

18 cytoldnes, transthyretin, amylin, lipoproteins, lympholdnes, lysozyme, a growth hormone (including human growth hormone), bovine growth hormone, growth hormone releasing factor, parathyroid hormone, thyroid stimulating hormone, growth factors, brain-derived neurotrophic growth factor, epidermal growth factor (EGF), fibroblast growth factor (such as D FGF and D FGF), insulin-like growth factor-I and -11, des(1-3)-IGF-I (brain IGF-I), insulin-like growth factor binding proteins, nerve growth factor (such as NGF-D), platelet derived growth factor (PDGF), vascular endothelial growth factor (VEGF), receptors for growth hormones or growth factors, transforming growth factor (TGF) (such as TGF-D, TGF-D 1, TGF-D2, TGF-D3, TGF-D4 or TGF-D5), neurotrophic factors (such as neurotrophin-3, -4 ,-5, or -6), gelsolin, glucagon, kallikreins, mullerian-inhibiting substance, neurotrophic factors, p53, protein A or D, prorelaxin, relaxin A-chain, relaxin B-chain, rheumatoid factors, rhodopsin, a serum albumin (such as human serum albumin), inhibin, insulin, insulin chains, insulin A-chain, insulin D -chain, insulin receptor, proinsulin, luteinizing hormone, integrin, interleuldns (ILs) (such as IL-1 to IL-10, IL-12, IL-13), erythropoietin, thrombopoietin, fibrillin, follicle stimulating hormone, clotting factors (such as factor V1IIC, factor 1X, tissue factor, and von Willebrands factor), anticlotting factors (such as Protein C, atrial naturietic factor, lung surfactant), a plasminogen activator (such as human tissue plasminogen activator or urokinase), thrombin, tumor necrosis factor- D or D, D-ketoacid dehydrogenase, addressins, bone morphogenetic proteins (BMPs), collagen, colony stimulating factors (CSFs) (such as M-CSF, GM-CSF and G-CSF), decay accelerating factor, homing receptors, interferons (such as interferon-alpha, -gamma and -beta), keratin, osteoinductive factors, PRNP, regulatory proteins, superoxide clismutase, surface membrane proteins, transport proteins, T- cell receptors, antigens such as gpl 20(HIb) immuno toxins, atrial natriuretic peptide, seminal vesicle exocrine protein, D 2- microglobulin, PrP, precalcitonin, ataxin 1, ataxin 2, ataxin 3, ataxin 6, ataxin 7, huntingtin, androgen receptor, CREB-binding protein, gpl 20, p300, CREB, API, ras, NFAT, jun, fos, dentaorubral pallidoluysian atrophy-associated protein, a microbial protein (e.g., maltose binding protein, ABC
transporter, glutathione S transferase, thioredoxin, D -lactamase), green fluorescent protein, red fluorescent protein, an enzyme such as superoxide dismutase, asparaginase, arginase, arginine deaminase, adenosine deaminase, ribonuclease, catalase, uricase, bilirubin oxidase, trypsin, papain, alkaline phosphatase, beta-glucoronidase, purine nucleoside

19 phosphorylase or batroxobin, an opioid, e.g. endorphins, enkephalins or non-natural opioids, a hormone or neuropeptide, e.g. calcitonin, glucagon, gastrins, adreno-corticotropic hormone (ACTH), cholecystoldnins, lutenizing hormone, gonadotropin-releassing hormone, chorionic gonadotropin, corticotrophin-releasing factor, vasopressin, __ oxytocin, antidiuretic hormones, thyroid-stimulating hormone, thyrotropin-releasing hormone, relaxin, prolactin, peptide YY, neuropeptide Y, pancreastic polypeptide, leptin, CART (cocaine and amphetamine regulated transcript), a CART related peptide, perilipin, melanocortins (melanocyte-stimulating hormones) such as MC-4, melanin-concentrating hormones, natriuretic peptides, adrenomedullin, endothelin, secretin, amylin, vasoactive intestinal peptide (VIP), pituary adenylate cyclase activating polypeptide (PACAP), bombesin, bombesin- like peptides, thymosin, heparin-binding protein, soluble CD4, hypothalamic releasing factor and melanotonins or functional analogs thereof. In another embodiment of the invention the peptide, polypeptide or protein of interest may be a processing enzyme such as proteases (eg.
enterokinase, __ caspases trypsine like serine proteases), lipase, phospatase, glycosyl hydrolases (eg. mannosidases, xylosidases, fucosidases), ldnase, mono or dioxidase, peroxidase, transglutaminase, car-boxypeptidase, amidase, esterase, phosphatase and the like.
Preferred sources for such mammalian polypeptides include human, bovine, equine, porcine, lupine and rodent sources, with human proteins being particularly preferred.
__ The peptide, polypeptide or protein of interest of the present invention also encompasses variants of the aforementioned peptide, polypeptide or protein of interest.
These variants encompass, for example, peptide, polypeptide or protein that has the same activity as the aforementioned peptide, polypeptide or protein of interest and that comprises an amino acid sequence with, in the amino acid sequence of the aforementioned peptide, __ polypeptide or protein of interest, one or more deleted, substituted, inserted and/or added amino acids. Such peptide, polypeptide or protein can be exemplified by peptide, polypeptide or protein that has the same activity as the aforementioned peptide, polypeptide or protein of interest and that comprises an amino acid sequence with, in the amino acid sequence of the aforementioned peptide, polypeptide or protein of interest, __ one or more deleted, substituted, inserted and/or added amino acids. Two or more different types of modifications selected from deletion, substitution, insertion, and addition may be carried out concurrently.
The peptide, polypeptide or protein of interest of the present invention also encompasses "partial peptides or polypeptides" of the aforementioned peptide, polypeptide or protein 5 of interest. A partial peptide or polypeptide of the peptide, polypeptide or protein of interest can be exemplified by a partial peptide or polypeptide comprising an amino acid sequence in which a portion of the amino acid sequence of the aforementioned peptide, polypeptide or protein of interest runs uninterrupted, wherein the partial peptide or polypeptide preferably has the same activity as said peptide, polypeptide or protein of 10 interest. Such a partial peptide or polypeptide can be exemplified by an amino acid sequence comprising at least 20 and preferably at least 50 of the amino acid residues in the amino acid sequence of the aforementioned peptide, polypeptide or protein of interest.
This peptide or polypeptide preferably contains the amino acid sequence that corresponds to the region that is involved with the activity of the aforementioned peptide, polypeptide 15 or protein of interest. In addition, the partial peptide or polypeptide used in the present invention may also be a partial peptide or polypeptide as yielded by a modification of this peptide wherein 1 or a plurality of amino acid residues (for example, approximately 1 to

20, more preferably approximately 1 to 10, and even more preferably approximately 1 to 5) is deleted from, substituted in, inserted into, and/or added to its amino acid sequence.
20 The partial peptide or polypeptide used in the present invention can also be used as an antigen for antibody production.
In one embodiment of the invention, the peptide, polypeptide or protein of interest is selected from the group comprising Human growth hormone, human insulin, follicle-stimulating hormone, Factor VIII, Erythropoeietin, Granulocyte colony-stimulating factor, Alpha-glactosidase A, Alpha-L-iduronidase, N-actetylgalactosamine-4-sulfatase, Dornase alfa, Tisssue plasminogen activator, Glucocerebrosidase, Interferon, Insulin-like growth factor 1, bovine somatotropin, Porcine somatotropin, bovine chymosin, and envelop protein of the hepaptitis B virus.
The peptide, polypeptide or protein of interest also encompasses modified peptides, polypeptides or proteins that have underwent posttranslational and post-export

21 modifications in the periplasm such as cyclization, glycosylation, phophorylation, methylation, oxidation, dehydratation, proteolytic cleavage.
In one embodiment, the peptide, polypeptide or protein of interest is an enzyme for metabolizing a biomolecule in the extracellular medium (herein referred as "extracellular biomolecule"). In one embodiment, the extracellular biomolecule comprises a polysaccharide or a lipid. In one embodiment of the invention, the polysaccharide comprises alginate, pectin, cellulose, cellobiose, laminarin, or a mixture thereof. In one embodiment of the invention, the lipid comprises a fatty acid, a glycolipid, a betaine lipid, a glycerolipid, a phospholipid, a glycerolphospholipid, a sphingolipid, a sterol lipid, a prenol lipid, a saccharolipid, a polyketide, or a mixture thereof. In one embodiment of the invention, the peptide, polypeptide or protein of interest is an enzyme converting a polysaccharide to a monosaccharide, an oligosaccharide, or both. In one embodiment of the invention, the peptide, polypeptide or protein of interest is an enzyme converting a lipid to a fatty acid, a monosaccharide, or both. In one embodiment of the invention, the monosaccharide or oligosaccharide is oligoalginate, mannuronate, guluronate, mannitol, a-keto acid, 4-deoxy-L-erythro-hexoselulose uronate (DEHU), 2-keto-3-deoxy D-gluconate (KDG), glucose, glucuronate, galacturonate, galactose, xylose, arabinose, or mannose. In one embodiment of the invention, the fatty acid is 14:0, trans-14, 16:0, 16:1n-7, trans-16, 16:2n-6, 18:0, 18:1n-9, 18:2n-6, 18:3n-6, 18:3n-3, 18:4n-3, 20:0, 20:2n-6, 20:3n-6, 20:4n-3,20:4n-6, or 20:5n-3.
In one embodiment of the invention, the peptide, polypeptide or protein of interest is an enzyme converting an extracellular biomolecule to a commodity chemical. In one embodiment of the invention, the commodity chemical is ethanol, butanol, or biodiesel.
In one embodiment of the invention, the biodiesel is a fatty acid, a fatty acid ester, or a terpenoid.
In one embodiment, the peptide, polypeptide or protein of interest is a secreted protein.
In one embodiment, the peptide, polypeptide or protein of interest is naturally secreted, i.e. the nucleic acid sequence encoding the same naturally comprises a signal peptide leading to its secretion. In another embodiment, the peptide, polypeptide or protein of interest is artificially secreted, i.e. the nucleic acid sequence encoding the same is fused

22 to a signal peptide, thereby allowing its secretion. Examples of signal peptides that may be used in the context of the present invention include, but are not limited to, OmpA
signal peptide (SEQ ID NO: 17), DsbA signal peptide (SEQ ID NO: 18) or PhoA
signal peptide (SEQ ID NO: 19).
In another embodiment, the peptide, polypeptide or protein is a transmembrane protein.
In another embodiment, the peptide, polypeptide or protein is a protein which is toxic for the host cell.
In one embodiment, the host cell of the invention further comprises a nucleic acid sequence encoding a protein that induces the expression of the peptide, polypeptide or protein of interest, preferably placed in an expression system.
Examples of such nucleic acid sequences include, but are not limited to, the gene encoding the T7 RNA polymerase, T7 gene 1. In that case, the expression of the polymerase induces the expression of the peptide, polypeptide or protein of interest placed under the control of a T7 promoter.
Preferably, the nucleic acid sequence encoding a protein that induces the expression of the peptide, polypeptide or protein of interest is the T7 expression system.
The T7 expression system was described in U54,952,496, which is incorporated herein by reference. The T7 expression system comprises a DNA fragment from the T7 phage, containing the entire coding sequence for the T7 RNA polymerase (i.e. the T7 gene 1).
Any natural active promoter of the 77 gene 1 was removed and an inducible lacUV5 promoter was inserted ahead of the coding sequence. The lacUV5 promoter is induced by addition of IPTG to the culture medium.
In one embodiment, the host cell of the invention comprises a phage inserted within its genome. In one embodiment where the peptide, polypeptide or protein of interest is placed under the control of a T7 promoter, said genetically modified phage is as described in the PCT patent application W02013/004817 (incorporated herein by reference). In one embodiment, said genetically modified phage is further dermed as a phage wherein:
- an expression system is inserted (preferably a T7 expression system), and

23 - the S and/or the R genes are inactivated.
In one embodiment of the invention, the Int and/or Xis gene(s) of the phage is/are inactivated.
In one embodiment, the genetically modified phage is further del-filed as a phage wherein:
- an expression system is inserted (preferably a T7 expression system), - the S, R and/or the Q gene(s) is/are inactivated, and - the Int and/or Xis gene(s) is/are inactivated.
Examples of phages which can be used in the invention include, but are not limited to, the lambda (X) phage (Enterobacteria phage lambda, accession number NC_001416), lambda-like and lambdoid phages. Lambda phage, also known as coliphage lambda, is a virus that infects Escherichia coll. Lambda is a temperate bacteriophage.
Lambda-like phages form a family of bacteriophages and archaeal viruses which are characterized by long, non-contractile tails. Lambdoid phages are natural relatives of lambda phage. Most of them grow on E. coli, but a few come from other host cells, such as, for example, Salmonella typhhnurium. Examples of lambda-like and lambdoid phages which could be used in the present invention include, but are not limited to, coliphage 434, phi80, phi81, HK97, P21 and P22.
In one embodiment, the phage is lambda DE3 (accession number EU078592). The Lambda DE3 phage is a modified lambda phage D69, comprising the gene encoding the T7 RNA polymerase under the control of a lacUV5 promoter.
According to a preferred embodiment, the genetically modified phage is the phage Pll having the sequence SEQ ID NO: 14 and being (DE3) AS-C, Axis-eal0 (DE3 refers to lambda phage DE3 wherein the T7 RNA polymerase gene has been integrated within the sequence of int gene). Said modified phage Pll corresponds to the sequence NC_001416 wherein the coding sequences of genes S, R, Rz, Xis and lnt are deleted.
According to another embodiment, the genetically modified phage is P12 corresponding to the sequence NC_001416 wherein the coding sequences of genes Int and S are deleted (one example of P12 is the sequence DE3 AS, Axis-eal0).

24 According to another embodiment, the genetically modified phage is P13 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, R, Rz, Q, Xis and Int are deleted (one example of P13 is the sequence DE3 AS-C, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P14 corresponding to the sequence NC_001416 wherein the coding sequences of genes R and Int are deleted (one example of P14 is the sequence DE3 AR).
According to another embodiment, the genetically modified phage is P15 corresponding to the sequence NC_001416 wherein the coding sequences of genes Q and Int are deleted (one example of P15 is the sequence DE3 AQ).
According to another embodiment, the genetically modified phage is P16 corresponding to the sequence NC_001416 wherein the coding sequences of genes S and Xis are deleted (one example of P16 is the sequence NC_001416 AS, Axis-eal0).
According to another embodiment, the genetically modified phage is P17 corresponding to the sequence NC_001416 wherein the coding sequences of genes R and Xis are deleted (one example of P17 is the sequence NC_001416 AR, Axis-eal0).
According to another embodiment, the genetically modified phage is P18 corresponding to the sequence NC_001416 wherein the coding sequences of genes Q and Xis are deleted (one example of P18 is the sequence NC_001416 Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P19 corresponding to the sequence NC_001416 wherein the coding sequences of genes R, Rz and Int are deleted (one example of P19 is the sequence DE3 AR, ARz).
According to another embodiment, the genetically modified phage is P20 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, Rz and Int are deleted (one example of P20 is the sequence DE3 AS, ARz).
According to another embodiment, the genetically modified phage is P21 corresponding to the sequence NC_001416 wherein the coding sequences of genes Rz, Q and Int are deleted (one example of P21 is the sequence DE3 ARz, AQ).

According to another embodiment, the genetically modified phage is P22 corresponding to the sequence NC_001416 wherein the coding sequences of genes S. Q and Int are deleted (one example of P22 is the sequence DE3 AS, AQ).
According to another embodiment, the genetically modified phage is P23 corresponding 5 to the sequence NC_001416 wherein the coding sequences of genes R, Q, and Int are deleted (one example of P23 is the sequence DE3 AR, AQ).
According to another embodiment, the genetically modified phage is P24 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, R and Int are deleted (one example of P24 is the sequence DE3 AS, AR).
10 According to another embodiment, the genetically modified phage is P25 corresponding to the sequence NC_001416 wherein the coding sequences of genes R, Rz and Xis are deleted (one example of P25 is the sequence NC_001416 AR, Axis-ea10, ARz).
According to another embodiment, the genetically modified phage is P26 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, Rz and Xis are 15 deleted (one example of P26 is the sequence NC_001416 AS, Axis-ea10, ARz).
According to another embodiment, the genetically modified phage is P27 corresponding to the sequence NC_001416 wherein the coding sequences of genes Q, Rz and Xis are deleted (one example of P27 is the sequence NC_001416 ARz, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P28 corresponding 20 to the sequence NC_001416 wherein the coding sequences of genes S, Q and Xis are deleted (one example of P28 is the sequence NC_001416 AS, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P29 corresponding to the sequence NC_001416 wherein the coding sequences of genes R, Q and Xis are deleted (one example of P29 is the sequence NC_001416 AR, Axis-ea10, AQ).

25 According to another embodiment, the genetically modified phage is P30 corresponding to the sequence NC_001416 wherein the coding sequences of genes R, S and Xis are deleted (one example of P30 is the sequence NC_001416 AS, Axis-ea10, AR).

26 According to another embodiment, the genetically modified phage is P31 corresponding to the sequence NC_001416 wherein the coding sequences of genes R, Xis and Int are deleted (one example of P31 is the sequence DE3 AR, Axis-eal0).
According to another embodiment, the genetically modified phage is P32 corresponding to the sequence NC_001416 wherein the coding sequences of genes S. Xis and Int are deleted (one example of P32 is the sequence DE3 AS, Axis-eal0).
According to another embodiment, the genetically modified phage is P33 corresponding to the sequence NC_001416 wherein the coding sequences of genes Q, Xis and Int are deleted (one example of P33 is the sequence DE3 Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P34 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, R, Xis and Int are deleted (one example of P34 is the sequence DE3 AS, Axis-ea10, AR).
According to another embodiment, the genetically modified phage is P35 corresponding to the sequence NC_001416 wherein the coding sequences of genes R, Q, Xis and Int are deleted (one example of P35 is the sequence DE3 AR, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P36 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, Q, Xis and Int are deleted (one example of P36 is the sequence DE3 AS, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P37 corresponding to the sequence NC_001416 wherein the coding sequences of genes R, Rz, Xis and Int are deleted (one example of P37 is the sequence DE3 AR, Axis-ea10, ARz).
According to another embodiment, the genetically modified phage is P38 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, Rz, Xis and Int are deleted (one example of P38 is the sequence DE3 AS, Axis-ea10, ARz).
According to another embodiment, the genetically modified phage is P39 corresponding to the sequence NC_001416 wherein the coding sequences of genes Rz, Q, Xis and Int are deleted (one example of P39 is the sequence DE3 ARz, Axis-ea10, AQ).

27 According to another embodiment, the genetically modified phage is P40 corresponding to the sequence NC_001416 wherein the coding sequences of genes S. R, Q and Int are deleted (one example of P40 is the sequence DE3 AS, AR, AQ).
According to another embodiment, the genetically modified phage is P41 corresponds to the sequence NC_001416 wherein the coding sequences of genes S, R, Rz and hit are deleted (one example of P41 is the sequence DE3 AS-C).
According to another embodiment, the genetically modified phage is P42 corresponding to the sequence NC_001416 wherein the coding sequences of genes R, Rz, Q and Int are deleted (one example of P42 is the sequence DE3 AR, ARz, AQ).
According to another embodiment, the genetically modified phage is P43 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, Rz, Q and Int are deleted (one example of P43 is the sequence DE3 AS, ARz, AQ).
According to another embodiment, the genetically modified phage is P44 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, R, Q and Xis are deleted (one example of P44 is the sequence NC_001416 AS, AR, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P45 corresponds to the sequence NC_001416 wherein the coding sequences of genes S, R, Rz and Xis are deleted (one example of P45 is the sequence NC_001416 AS-C, Axis-eal0).
According to another embodiment, the genetically modified phage is P46 corresponding to the sequence NC_001416 wherein the coding sequences of genes R, Rz, Q and Xis are deleted (one example of P46 is the sequence NC_001416 AR, ARz, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P47 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, Rz, Q and Xis are deleted (one example of P47 is the sequence NC_001416 AS, ARz, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P48 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, R, Q, Xis and Int are deleted (one example of P48 is the sequence DE3 AS, AR, Axis-ea10, AQ).

28 According to another embodiment, the genetically modified phage is P49 corresponding to the sequence NC_001416 wherein the coding sequences of genes S. Rz, Q, Xis and Int are deleted (one example of P49 is the sequence DE3 AS, ARz, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P50 corresponding to the sequence NC_001416 wherein the coding sequences of genes R, Rz, Q, Xis and Int are deleted (one example of P50 is the sequence DE3 AR, ARz, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P51 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, R, Q, Rz and Int are deleted (one example of P51 is the sequence DE3 AS-C, AQ).
According to another embodiment, the genetically modified phage is P52 corresponding to the sequence NC_001416 wherein the coding sequences of genes S, R, Rz, Q
and Xis are deleted (one example of P52 is the sequence NC_001416 AS-C, Axis-ea10, AQ).
According to another embodiment, the genetically modified phage is P53 corresponding to the sequence NC_001415 wherein the coding sequences of genes Q, Xis and Int are deleted (one example of P53 is the sequence DE3 Axis-ea10, AQ).
The present invention also relates to a process for producing a peptide, polypeptide or protein of interest, wherein said method comprises:
- cultivating a host cell of the invention, comprising:
- at least two copies of a nucleic acid sequence encoding a poison protein, preferably CcdB, - at least one copy of a nucleic acid sequence encoding the antidote protein to the poison protein, preferably CcdA, and - at least one copy of the nucleic acid sequence encoding the peptide, polypeptide or protein of interest, and - recovering the peptide, polypeptide or protein of interest.
In one embodiment, the at least two copies of a nucleic acid sequence encoding a poison protein are inserted within the genome of the host cell.

29 In one embodiment, the at least one copy of a nucleic acid sequence encoding the antidote protein to the poison protein and the at least one copy of the nucleic acid sequence encoding the peptide, polypeptide or protein of interest are carried by the same vector.
Therefore, only host cells containing the vector are propagated, while host cells without the vector die due to the toxic effect of the poison protein.
In one embodiment of the invention, the nucleic acid sequence encoding the peptide, polypeptide or protein of interest is comprised within the expression system.
According to this embodiment, the production of the peptide, polypeptide or protein of interest is direct, i.e. it results from the expression of the gene of the expression system, for example by culture in a medium wherein the promoter comprised in the expression system is induced. For example, in one embodiment, the nucleic acid sequence encoding the peptide, polypeptide or protein of interest is placed under the control of a pLac or pTRC
promoter, and the expression of the peptide, polypeptide or protein of interest is induced when IPTG is added to the culture medium. Moreover, in another embodiment, the nucleic acid sequence encoding the peptide, polypeptide or protein of interest is placed under the control of a Para promoter, and the expression of the peptide, polypeptide or protein of interest is induced when arabinose is added to the culture medium.
In another embodiment, the host cell of the invention further comprises another expression system comprising the nucleic acid sequence of the T7 RNA
polymerase under the control of a lac promoter, preferably the lacUV5 promoter, inserted in its genome.
According to this embodiment, the process for producing the peptide, polypeptide or protein of interest comprises the transformation of the host cell with a vector, preferably a plasmid, comprising the nucleic acid sequence of the peptide, polypeptide or protein of interest under the control of the T7 promoter and the nucleic acid sequence encoding the antidote protein. The expression from the T7 promoter is under the control of polymerase, which is stringently specific for the T7 promoter, i.e. the T7 promoter can only be utilized by the RNA polymerase of bacteriophage T7. When lPTG is added to the culture medium, T7 RNA polymerase is expressed by transcription from the lac promoter which will allows the expression of the peptide, polypeptide or protein of interest.

According to one embodiment, the peptide, polypeptide or protein of interest is secreted by the host cell in the fermentation broth. According to this embodiment, the peptide, polypeptide or protein of interest may be easily recovered from the fermentation broth using methods well-known in the art.
5 According to another embodiment, the peptide, polypeptide or protein of interest is not secreted by the host cell in the fermentation broth. Methods for recovering an intracellular or periplasmic peptide, polypeptide or protein of interest are also well-known in the art.
Examples of such methods include, but are not limited to, the use of trichloroacetic acid (TCA) or cracking buffer containing sodium dodecyl sulfate (SDS) to recover total 10 proteins in denaturing conditions or the use of sonication, French press or equivalent to disrupt bacteria under pressure in order to recover total cytoplasmic proteins in native (not denaturing) conditions. Next, the peptide, polypeptide or protein of interest can be purified using specific methods including but not limited to the use of affinity or ion exchange columns.
15 The present invention also refers to a method for preparing a host cell as hereinabove described.
The method for preparing a host cell of the invention comprises inserting at least two copies, preferably two copies, of the nucleic acid sequence encoding the poison protein within the genome of the host cell. Preferably, said poison protein is CcdB, more 20 preferably said the nucleic acid sequence encoding the poison protein is SEQ ID NO: 1.
In one embodiment, the method for preparing a host cell of the invention comprises a step of deletion, wherein nucleic acid sequences of the host cell are deleted.
Methods for inserting or deleting the sequence of a gene are well known by the skilled artisan. The more efficient method is the homologous recombination method mediated 25 by the lambda Red-encoded genes or the recE and recT genes from the prophage Rac.
This method was well described by several researchers including Datsenko and Wanner (PNAS 97-12, 6640-6645, 2000) and Stewart et al. (W00104288). PCR products are generated using primers with 20- to 60-nt extensions that are homologous to regions adjacent to the gene to be deleted or to the region wherein a nucleic acid sequence will be inserted. Since only a small amount of bacteria will effectively recombine the fragment of interest, it is necessary to have a strong selection marker to select it.
Antibiotic markers can be used to select the recombinants: the modified primers are used to amplify an antibiotic resistance gene. After transformation and activation of the recombination genes, recombinant bacteria are selected on medium containing the appropriate antibiotic.
In this case, the targeted gene or region is replaced by an antibiotic resistance gene. In order to use the same strategy for the next deletion, it is necessary to remove this antibiotic resistance gene during a second step. As described in Datsenko and Wanner, it is possible to use antibiotic resistance gene that are flanked by FRT (FLP recognition target) sites.
The resistance genes are then eliminated by using a helper plasmid encoding the FLP
recombinase. The antibiotic resistance gene is removed by this site-specific recombinase but this method leaves traces: one site-specific recombination site is still present after removal of the antibiotic resistance gene.
To avoid the presence of this site, more preferably, the method of the invention uses galK
as a marker gene. The principle of the galK selection is described in Warming et al.
(Nucleic acid research, 2005,33(4)). This method uses galK as a positive selection marker (growth on minimal medium containing galactose) during the first recombination (insertion). The removal of this marker is performed during a second homologous recombination step. During this step, galK is used as a negative selection marker on minimal medium containing 2-deoxy-galactose (DOG). The galK gene product, galactoldnase, catalyzes the first step in the galactose degradation pathway.
Galactoldnase also efficiently catalyzes the phosphorylation of the DOG galactose analog.
The product of this reaction cannot be further metabolized, leading to the accumulation of a toxic molecule (2-deoxy-galactose- 1 -phosphate). The advantage of this method is to avoid the presence of specific recombination site after deletion of the targeted gene and removal of the selective marker.
In one embodiment, the method of the invention further comprises a step of infection of the host cell by a genetically modified phage as defined hereinabove. In one embodiment of the invention, said infection step includes the use of a helper phage. In the meaning of the present invention, the term "helper phage" refers to a phage used to complement a deletion or an inactivation of another phage. The helper phage will provide the missing functions to said another phage to be able to infect bacteria or to prepare phage stock.
Usually, the helper phage cannot form a lysogen by itself because it is cI
minus (it has no repressor and is thus virulent).
Processes for infecting a host cell with a phage using a helper phage are well known in the art. In one embodiment, a first step is the preparation of the lysates and a second one is the lysogenization. Briefly, the bacterial lysates of the helper phage may be prepared using standard methods as described in "Molecular cloning: a laboratory manual", Sambrook et al. (2001, ISBN 978-087969577-4) or in "Large- and Small-Scale Preparation of Bacteriophage lambda lysate and DNA", Su et al., BioTechniques 25:44-46 (July 1998). Preparation of the phage of interest may be done using the same principle, after phage induction (most often using UV irradiation or any situation where a lysogen undergoes DNA damage or the SOS response of the host or Cro production) in order to launch the lytic cycle and using a helper phage to provide the missing functions. An alternative to the helper phage is the use of a plasmid encoding the missing functions.
Next, the phage lysates may be mixed with the targeted bacteria and plated on LB plates in order to get lysogens (as described in lambda DE3 lysogenization kit from Novagen, User Protocol TB031 or an alternative method is described in Studier and Moffat, Journal of Molecular Biology, 1986, 189:113-130). A selection phage can be used to select specifically bacteria containing the phage of interest. This selection phage is a virulent phage having the same immunity as the phage of interest. Consequently, the selection phage is unable to form plaques or to kill bacteria lysogens for the phage of interest because this phage produces the cI repressor (also called C2 in DE3 lambda phage).
In one embodiment, the method of the invention further comprises transforming the host cell of the invention with a vector, preferably a plasmid as described hereinabove, carrying (1) a nucleic acid sequence encoding the antidote to the poison protein and (2) a nucleic acid sequence encoding the peptide, polypeptide or protein of interest. Methods for transforming a cell with a vector are well-known in the art and include, but are not limited to, chemical transformation and electroporation.

The method for producing a peptide, polypeptide or protein of interest using the host cell of the invention thus presents the following advantages:
- it may be applied to any peptides, polypeptides or proteins, even to peptides, polypeptides or proteins which are toxic to the host cell;
- it may be applied to any host cell, as the insertion of the ccdB gene within the genome of a host cell is easy for the skilled artisan;
- as shown in the examples below, it results in an increased production yield of the peptide, polypeptide or protein of interest;
- as shown in the examples below, it allows increasing stability of the vector comprising the nucleic acid sequence encoding the peptide, polypeptide or protein of interest, especially when the peptide, polypeptide or protein of interest is toxic for the host cell; and - in some embodiments, it does not comprise using antibiotics, and both the host cell and the vector are free of antibiotic-resistance genes.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 is a map of the pVHH6 vector.
Figure 2 is a photograph showing the PCR screening on the W3110ccdB strain after 24h of induction of pVHH6AKan. Culture after induction has been streaked on plates without antibiotics and PCR done on 20 colonies. PCR results over VHH6 (A), ccdB (B) and strain specific (C) genes are shown. Arrows on the left show expected size of the PCR
products. Head arrows show ccdB alterations.
Figure 3 is a photograph showing production comparison of W3110 pVHH6, W3110ccdB pVHH6AKan and W31102ccdB pVHH6AKan strains on SDS-PAGE.
Differences are observed between non-induced (NI) and induced (I) samples, but especially between induced samples (rectangle).
Figure 4 is a map of the pTraG vector.

Figure 5 is a photograph showing production comparison of W3110ccdB pTraG and W31102ccdB pTraG strains on SDS-PAGE. Differences are observed between non-induced (NI) and induced (I) samples, but not between induced samples.
Figure 6 is a combination of two photographs showing production comparison of pD1.3, W3110ccdB pD1.3ccdAAAmp and W31102ccdB pD1.3ccdAAAmp strains on SDS-PAGE by Coomassie blue staining (left) or by western blot (right).
Differences are well observed by western blot between non-induced (NI) and induced (I) samples.
Figure 7 is a photograph showing screening PCR on the SE3 (BL21(DE3)2ccdB) strain after 24h of induction of pStaby1.2 TraG. Culture after induction has been streaked on Petri dishes and PCR done on 22 colonies. PCR results over TraG (A), ccdA (B), the first ccdB (C) and the second ccdB (D) genes are shown. Arrows on the left show expected size of the PCR products.
Figure 8 is a photograph of a SDS-PAGE gel showing production comparison of pStabyTraG and SE4 (BL21(DE3)2ccdB 2gyrA) pStabyTraG strains. Differences are observed between non-induced and induced (I) samples.
Figure 9 is a photograph showing production comparison of W3110ccdB pVHH6AKan, W31102ccdB pVHH6AKan and W31102ccdB 2gyrA pVHH6AKan strains on SDS-PAGE. Differences are observed between non-induced (NI) and induced (I) samples, but especially between induced samples.
EXAMPLES
The present invention is further illustrated by the following examples.
Example 1: Duplication of the ccdB gene increases yield production In order to test the stability brought by the Staby@ technology (technology based on the insertion of one copy of the ccdB gene within the host chromosome, Delphi Genetics, StabyExpress@ products, patent W09958652), we produced a secreted protein (a variable heavy chain antibody fragment named VHH6) which seems to have also toxic properties in E. coli. The W3110 E. coli strain (genetic background: E. coli K12, genotype: F
lambda- 1NV(rrnD-rrnE) rph-1) was used as producer host and expression was realized with pTrc promoter on a pBR322 vector (pVHH6; Fig. 1). The production of the strain using classical antibiotic stabilisation system (here the kanamycin resistance) was 5 compared to the production of the W3110ccdB strain which has been modified to use the Staby@ technology (ccdB gene being introduced inside the strain chromosome).
Vectors used for this comparison both contained the ccdA gene and the vector used in W3110ccdB
was deleted for the kanamycin resistance gene (pVHH6AKan).
Strains were grown at 30 C in LB medium (10m1) with or without antibiotics and induced 10 (0,5mM 1PTG) for 4h when OD600 reaches 0.5. No production was visible when analysed by SDS-PAGE despite 100% of plasmid stability in both selection systems under these gentle induction conditions. Stronger induction condition has been applied as followed (here after continuous fermentation based on perfusion). Strains were grown in 500m1 fermenters and OD600 was maintained constant at 0.3 by perfusion: addition of fresh 15 medium (up to 51) and removing of the overflow to maintain always OM to 0.3 in the fermenter. After 51 of perfusion, the OM increased and induction was finally realized for 24h when cultures reached an OD600 of 0.5. When antibiotic selection system is used to stabilize the expression vector, the final plasmid stability was determined by plating bacteria on plates containing or not the appropriate antibiotics (according to the resistance 20 gene encoded by the plasmid). Plates without antibiotics is used to determine the total number of bacteria and plates containing antibiotics are used to determine the number of bacteria still containing the plasmid.
The plasmid stability was evaluated to about 0% for the bacteria using the antibiotics (no colony on plates containing antibiotics).
25 When using the Staby@ system based on ccdA and ccdB, no antibiotic resistance gene is present. Consequently, the presence or absence of the plasmid is tested by PCR. Since we are using the Staby@ system, we hypothesize that due to the presence of the selection ccdB gene in the chromosome bacteria and the ccdA antidote gene in the plasmid, all bacteria growing on plates should contain the plasmid. Unexpectedly, we observed a

30 stability of about 65% by PCR experiments on colonies (Fig. 2B). Indeed, even if the stability is drastically increased, this is the first time that the Staby technology does not give 100% of stability.
Moreover, among the 65% of retention, the gene of interest (VHH6 gene) was mutated by partial or full deletions in all plasmid tested (Fig. 2A). Bands of ccdB
PCR which are higher than expected were extracted and sequenced. Sequencing results show that the ccdB gene is inactivated by small insertion sequence (insAB) insertion (SEQ ID
NO: 16).
According to the strain specific PCR results (Fig. 2C), unamplified band over ccdB gene could be due to large deletion or insertion in the ccdB region. This experiment has been replicated and equivalent results were observed.
Therefore, the Staby technology has to be improved in order to avoid this kind of stabilisation system inactivation. A second copy of the poison gene has been inserted between the yjjK and sit genes (resulting in W31102ccdB strain). These genes have been chosen because they stand at the opposite side of the E. coli chromosome regarding to the first ccdB gene location and because the intergenic space seems to be non-coding.
In order to compare the production in W3110, W3110ccdB and W31102ccdB strains, the following protocol has been applied. Single colonies were grown overnight in 10 ml LB
medium at 37 C. The following day, 1 I from each culture was used for inoculation of 10 ml fresh LB medium. Cultures were grown overnight at 37 C. This step was repeated once again to perform additional generations (as realized in the fermenter).
The following day, 10 pi from each culture was used for inoculation of 10 ml fresh LB
medium. Cultures were grown at 30 C and induced at OD600 0.5 for 24h. Kanamycin was used all along these steps only for the W3110 strain. Despite the use of this antibiotic to maintain the expression plasmid, its stability was about 0% after induction. On the other side, the plasmid stability was about 100% for both Staby systems (i.e. with 1 or 2 ccdB genes) under this more gentle protocol: every PCR done over ccdB genes and plasmid ccdA
genes were at the expected size. Only some VHH6 genes were mutated (27%) for the W31102ccdB culture.
Productions were also compared by Coomassie blue SDS-PAGE (Fig. 3).
Unexpectedly, the yield of production of the W31102ccdB strain is much higher than with other strains (VHH6 = 27kDa). This kind of yield increase has already been reported by using the Staby@ technology but with only one copy of the ccdB gene (Peubez et al., Microbial Cell Factories 2010,9:65; Sodoyer et al., In: Antibiotic Resistant Bacteria -A Continuous Challenge in the New Millennium 2012). Here, the yield increase surprisingly happens only when two ccdB copies are present despite the fact that the plasmid is stable and present in the bacteria with one copy of the ccdB gene.
Example 2: Application of the Staby 2ccdB technology to produce other proteins In order to test if the use of the strains with two copies of the ccdB gene is possible for all proteins (including proteins already well expressed with only one copy of ccdB), we tested the production of the TraG protein of 70kDa (Szpirer et al., 2000, Mol.
Microbiol.
37(6) 1283-1292). W3110(DE3)ccdB and W3110(DE3)2ccdB strains with the plasmid coding for the antidote CcdA and TraG (pTraG) were constructed (Fig. 4). The traG gene was under the control of the T7 promoter. These strains and plasmid use the Staby@
technology to stabilize the plasmids. The first observation was that in the presence of the plasmid encoding ccdA, the presence of two copies of the ccdB gene does not affect the viability nor the growth of the W3110 strain: one copy of the ccdA gene in the plasmid is enough to get full viability.
The production of TraG under the classical Staby@ system (1 ccdB gene) is equivalent to the one with the new Staby@ system (2 ccdB genes) (Fig. 5). However the production with 1 ccdB gives already very good yield, so an increase of this yield is maybe not possible, even by applying the new Staby@ system.
In order to confirm our previous results, another production has been realized. An antibody fragment has been chosen as a model (D1.3 = 24kDa for heavy and light chains or 50kDa for the associated chains) and expressed using a Para promoter (arabinose induction). The vector (pBAD24) has been adapted to the strains containing the ccdB
gene(s) (insertion of ccdA and deletion of the ampicillin resistance gene).

As shown in Figure 6, the production yield is strongly increased by using 1 or 2 copies of the ccdB genes but we did not observe a difference between the two strains (including growth and speed).
In conclusion, the use of the strain with 2 copies of ccdB may be used for the production of any protein. However, the yield advantage depends on each protein and when the yield is already very high when using only 1 copy of ccdB, the strain encoding 2 copies gives the same results. Since, as industrial producers and researchers never know whether a protein will be difficult to produce or not when they are starting their experiments, it is an advantage to be able to use the same strain for classical and difficult-to-produce proteins and to start immediately with the strain designed for the hardest proteins.
Example 3: Application of the 2ccdB technology to produce proteins in other host strains E. coli strains are divided in K12 and B types. As the W3110 strain is a K12 strain, the applicability of the 2ccdB technology to the B strains has been tested. Thus two copies of ccdB have been introduced in the genome of an E. coli B (BL21(DE3) type ¨
Genetic background: E. coli B - genotype F¨ ompT gal dcm ion hsdSB(rB- mB-) X(DE3 [lad lacUV5-T7 gene 1])) to give the SE3 strain. Then, perfused fermentation has been realized to produce the TraG protein after several generations. Stability of the plasmids after 24h of induction is of 100% (Fig. 7). Every ccdB genes are intact under these conditions of expression.
Similar results were obtained in another K12 derived strain having the following genotype: F- endA 1 glnV44 thi-1 recA 1 relA 1 gyrA96 deoR nupG 1080dlacZAM15 A(lacZYA-argF)U169, hsdR17(rK- mK+), dcm::FRTccdB yjjK::FRTccdB. This strain was derived from DH5 alpha in which 2 copies of ccdB gene were inserted as described above.
These results show that it is possible to construct different E. coli strains containing the insertion of 2 copies of the ccdB gene. The presence of these 2 copies is not deleterious.

Example 4: Duplication of the gyrA gene As described earlier, the Staby technology is based on the toxicity of the CcdB protein for E. coli. As the CcdA antidote inhibits the expression of the ccdB gene, no selection pressure is applied against this toxin gene. It is so unlikely to find some E.
coli resistant strains to CcdB. However, this kind of resistant has been artificially isolated by applying several round of growth in presence of mutagen agent (Bernard et al., J Mol Biol. 1992, 226:735). This resistant strain presents a mutation in the gyrA gene (the target of CcdB) changing the arginine 462 in a cysteine.
During the expression of our toxic protein model VHH6 (see Example 1), strains had been isolated with wild-type ccdB gene but without the plasmid (and so the ccdA gene).
Thus, these strains were resistant to CcdB. The gyrA gene of these strains has been sequenced to confirm the arginine 462 modification in another already undescribed amino acid (serine).
By duplicating the ccdB gene as mentioned in this patent, we guess that the pressure of selection on the gyrA gene will increase and generate more easily CcdB
resistant strains.
In order to avoid those mutants, the duplication of the gyrA gene has also been realized.
This new copy of gyrA has been integrated in the same time that the second copy of ccdB
(in the same insert) between the yjjK gene and the sit gene in the SE2 strain by transduction of a DNA fragment containing the gyrA, kanamycin and ccdB genes surrounded by homologous regions used as recombination arms. The recombination event was selected using the kanamycin resistance gene. This gene was surrounded by two FRT
sites allowing its removal using the FLP recombinase. This construction has been successfully done and the viability of the strain has not changed. TraG
expression abilities were tested as in Example 2 and performed with the derivative of SE2 strain called SE4 (BL21(DE3)2ccdB 2gyrA).
As shown on Figure 8, the production of TraG under the classical Staby system (1 ccdB
gene) is equivalent to the one with the new Staby system (2 ccdB genes + 1 additional copy of gyrA). However the production with 1 ccdB gives already very good yield, so an increase of this yield is maybe not possible, even by applying the new Staby system.

We can observe that the addition of 2 ccdB and 1 additional copy of gyrA (1 copy is already present) do not reduce the amount of protein due to higher toxicity of the CcdB
poison protein.
Similar genetic modifications were applied to E. coli W3110 derivative (strain called 5 KW4: W31102ccdB 2gyrA). This strain has been compared to other W3110 strains for the production of the toxic protein model VHH6. After production, each strain was tested regarding the presence of the plasmid and 100% of retention was observed in all strains.
As shown in Figure 9, the production of VHH6 is better when 2 copies of the ccdB gene are present (Figure 9). Moreover, as mentioned for the SE4 strain, the addition of 2 ccdB
10 and 1 additional copy of gyrA (1 copy is already present) does not reduce the amount of protein produced and does not affect growth ability.
Example 5: Use of ccdB from E. coli 0157 In order to validate that copies of the ccdB gene from another origin than the F plasmid may be used in the 2ccdB system, we constructed a strain containing a copy of ccdB from 15 E. coli 0157:H7. The ccdB gene from E. coli 0157:H7 has been chosen (1) because it has only 36% of homologies with the ccdB gene from the F plasmid (at amino acids level) and (2) because it is inhibited by the same CcdA protein from the F plasmid.
At first, a DNA fragment has been constructed containing the ccd/30157 gene and a kanamycin resistance gene surrounded by 2 FRT sites allowing the excision of the 20 resistance gene after insertion of the DNA fragment into the bacterial genome. This construction was ligated in a plasmid and transformed in B462 (a strain resistant to CcdB
due to the presence of the gyrA462 mutation). In a second time, this fragment has been inserted between the yijK and sit genes of E. coli W3110 containing a plasmid encoding ccdA under the control of the Pbad promoter to generate the new KW5 strain (W3110 25 yjjK ::FRTccdBoi57) after removal of the resistance gene. This strain grows as the WT
strain and the ccdBoir gene kills bacteria as expected when the Pbad promoter is repressed (in the absence of arabinose). These results show that it is possible to construct a strain using the ccdB0157 gene. The CcdB0157 protein product is counteracted by the F
plasmid CcdA and usable in the present invention.

Example 6: Further modifications of the host cell of the invention We next tested the applicability of the 2ccdB system to a bacterial strain comprising a genetically modified phage comprising an expression system, and wherein the Q, S, R
and Rz genes are deleted. The deletion of these genes prevents lysis of the host cell when accidental reactivation of the phage occurs.
A DNA fragment containing the kanamycin resistance gene with 2 FRT sites and surrounded by the region bordering the Q, S. R and Rz genes in order to allow their deletion by homologous recombination (as previously described in W02013004817) was transduced in the SE2 strain [BL21(DE3) dcm::FRTccdB] to generate the strain BL21(DE3) dcm::FRTccdB QSRRz::FRT after removal of the kanamycin resistance gene. In a second time, another fragment encoding yjjK::FRTkanFRTccdB has been transduced in the BL21(DE3) dcm::FRTccdB QSRRz::FRT to generate the SE5 [BL21(DE3) dcm::FRTccdB yjjK::FRTccdB QSRRz::FRT] after removal of the kanamycin resistance gene. This final strain contains 2 ccdB genes as described above and the deletion of the Q, S, R, Rz genes avoiding bacterial lysis.
The SE5 strain has been compared to its unmodified-QSSRz homologous strain (SE3).
There are growing both at the same rate of the SE1 strain and the production of a control protein (TraG) gave also the same results.

Claims (15)

1. Host cell for producing a recombinant peptide, polypeptide or protein of interest, wherein said host cell comprises at least 2 copies of a nucleic acid sequence encoding a poison protein, provided that said at least 2 copies are in a different replicon than a nucleic acid sequence encoding the antidote protein to the poison protein.

2. Host cell according to claim 1, wherein said poison protein is CcdB, encoded by SEQ ID NO: 1 or any nucleic acid sequence having at least 75% identity with SEQ
ID NO: 1.

3. Host cell according to claim 1 or claim 2, wherein said cell further comprises at least one copy of a nucleic acid sequence encoding the antidote protein to the poison protein.

4. Host cell according to claim 3, wherein said antidote protein is CcdA, encoded by SEQ ID NO: 13 or any nucleic acid sequence having at least 75% identity with SEQ
ID NO: 13.

5. Host cell according to claim 3 or claim 4, wherein said nucleic acid sequence encoding the antidote protein is carried by a plasmid further comprising an expression system wherein the nucleic acid sequence encoding a recombinant peptide, polypeptide or protein of interest is or may be inserted.

6. Host cell according to claim 5, wherein said expression system comprises a promoter selected from the group comprising a T7 promoter, Ptrc, Para and Plac.

7. Host cell according to any one of claims 5 or 6, wherein said expression system comprises a T7 promoter and the host cell further comprises a T7 RNA
polymerase gene, preferably inserted in its genome, more preferably in a T7 expression system.

8. Host cell according to any one of claims 5 to 7, wherein said expression system comprises a T7 promoter and the host cell further comprises a genetically modified phage inserted within its genome, preferably said phage is defined as a phage wherein:
- a T7 expression system is inserted, - the S, R, and/or the Q genes are inactivated, and - the Int and/or Xis gene are inactivated.

9. Host cell according to any one of claims 1 to 8, being a bacterium, preferably a gram negative bacterium, more preferably an Enterobacteriacea, and even more preferably E. coli.

10. Host cell according to any one of claims 1 to 9, further comprising inactivation of at least one of the genes tonA, galK, araB, araA, Ion, ompT, rcsA, hsdR, mrr, endA
and recA.

11. Host cell according to any one of claims 1 to 10, further comprising at least one additional copy of the gyrA gene.

12. A kit comprising a host cell according to any one of claims 1 to 11 and a vector comprising at least one copy of the nucleic acid sequence encoding the antidote protein and at least one copy of an expression system wherein the nucleic acid sequence encoding a recombinant peptide, polypeptide or protein of interest is or may be inserted.

13. The kit according to claim 12, wherein the nucleic acid sequence encoding a peptide, polypeptide or protein of interest is under the control of a promoter selected from the group comprising a T7 promoter, Ptrc, Para and Plac.

14. A method for producing a recombinant peptide, polypeptide or protein of interest, wherein said method comprises cultivating the host cell of any one of claims 1 to 11 and recovering the peptide, polypeptide or protein of interest.

15. A method for producing a recombinant peptide, polypeptide or protein of interest according to claim 14, wherein the recombinant protein is a secreted protein, a transmembrane protein or a protein which is toxic for the bacterial strain.